Biochemical and biophysical characterization of the RNA-dependent RNA polymerase from Chikungunya virus and discovery of a novel inhibitor as a potential drug candidate

Chikungunya virus (CHIKV) is the causative agent of Chikungunya fever, an acute febrile and arthritogenic illness with no effective treatments available. The development of effective therapeutics strategies could be signicantly accelerated with detailed knowledge of the molecular components behind CHIKV replication. However, drug discovery is hindered by our incomplete understanding of their main components. The RNA-polymerase RNA-dependent (nsP4-CHIKV) is considered the key enzyme of the CHIKV replication complex and a suitable target for antiviral therapy. Herein, the nsP4-CHIKV was extensively characterized through experimental and computational biophysical methods. In the search for new molecules against CHIKV, a compound designated LabMol-309 was identied and mapped to bind to the nsp4-CHIKV active site. The inhibitory activity of LabMol-309 was evaluated in cellular-based antiviral assays using a CHIKV replicon system and a reporter virus. In conclusion, this study highlights biophysical features of nsP4-CHIKV and identied a new compound as a promising antiviral agent against CHIKV infection. of cys-labelled with An initial test with check a serial dilution of the from to


Introduction
The Chikungunya virus (CHIKV) belongs to the Togaviridae family and is the causative agent of Chikungunya fever. The main transmission route occurs through the bite of infected female mosquitoes of the Aedes sp. Genus. After CHIKV infection, the proportion of individuals who develop clinical and debilitating symptoms is considered the highest compared to other arboviruses, with an average of 80% of symptomatic cases (1,2). The control of the mosquito vector remains the best prophylaxis since there are no licensed vaccines or e cient antivirals available (3). In this scenario, the infection caused by CHIKV has a high social impact and constitutes a serious public health issue (3).
Chikungunya fever presents an acute phase characterized by high fever, arthralgia, myalgia, headaches, edema, periorbital pain and cutaneous rash (4,5). Later, some patients progress to the so-called chronic phase, mainly characterized by persistent arthralgia and musculoskeletal pain for months and even years (4,(6)(7)(8). Due to the lack of speci c antiviral drugs, the infection is treated symptomatically, using analgesics and antipyretics in the acute phase and also with non-steroidal anti-in ammatory drugs (NSAIDs) and corticosteroids in the chronic phase (5).
CHIKV is a spherical, enveloped, and positive single-stranded RNA virus. As a member of the Alphavirus genus, its genome has approximately 12 kb and codes for two distinct polyproteins: structural and nonstructural (9). The rst one is cleaved and gives rise to ve structural proteins -E1, E2, E3, C and 6kwhich are part of the structure and viral assembly (9). The envelope proteins, specially E2 and E1, are responsible for virus anchoring, receptor interaction and membrane fusion, promoting virus entry in the host cell (10,11). Due to its location on the viral surface, envelope proteins are targets of the humoral immune response and thus become targets for the vaccine's development against CHIKV (12).
On the other hand, the non-structural polyprotein is cleaved in four non-structural proteins (nsP1, nsP2, nsP3 and nsP4) that form the viral replication complex and also have some functions in the infection process, such as interaction with host factors (9,13,14). The nsP1 has methyltransferase and guanylyltransferase activities, promoting the capping of viral RNA. In addition, this protein also anchors the replication complex in the cell membrane (15). The nsP2 is a multifunctional protein with NTPase and helicase activities in the N-terminal region. Its C-terminus has a cysteine-protease activity, responsible for processing the non-structural polyprotein (15). The nsP3 is an accessory protein that recruits host cell factors that participate and optimize the replicative process (16). Finally, nsP4 is the RNA-dependent RNA polymerase (RdRp), which is considered the key enzyme of the CHIKV replication complex and acts principally by promoting the synthesis and elongation of viral RNA (17).
Therefore, due to their vital participation in the viral life cycle, non-structural proteins emerge as potential targets for the antiviral drugs' development, aiming to interrupt the replication process and, consequently, the viral elimination (18). Among these proteins, nsP4 is an attractive target due to its central role in viral genome replication, transcription and some acting in genome repair (17,19). To date, no threedimensional structure of the nsP4 of CHIKV (nsP4-CHIKV) have been reported. This fact generates limitations and challenges when the nal goal is applying a structure-based drug design against this target.
Due to the lack of high-resolution structural information of the nsP4-CHIKV and the need to search for new drugs to treat the infection caused by this virus, this work presents a detailed biophysical characterization of this protein. Size exclusion chromatography coupled with multi-angle light scattering was used to infer the oligomeric state of the nsp4-CHIKV protein in solution. A high prevalence of ordered helical secondary structures was observed by circular dichroism, which also showed that the nsP4-CHIKV unfolds under a cooperative transition during thermal denaturation. The thermal denaturation was further studied using differential scanning calorimetry, indicating a kinetically controlled process.
Moreover, in the search for ligands towards the development of new inhibitors, 12 compounds were screened using differential scanning uorimetry experiments. One compound, LabMol-309, showed a signi cant effect on nsP4-CHIKV stability and was submitted to MicroScale Thermophoresis experiments to con rm its interaction with the protein. Furthermore, this interaction was also analyzed by nuclear magnetic resonance, monitoring chemical shift perturbation. To complete, in silico studies of molecular docking and molecular dynamics simulations were performed with the nsP4-CHIKV 3D model and the compound LabMol-309, to suggest the possible binding mode and maintenance of LabMol-309 at the protein active site.
Furthermore, the inhibitory activity of LabMol-309 was studied in cellular-based antiviral assays using a CHIKV replicon system and a reporter virus, and the results showed that this molecule caused inhibition in these cellular assays and has the potential to be further evaluated as a CHIKV inhibitor.
In this way, this work provides novel structural features of nsP4-CHIKV and identi ed a new compound that interacts with this protein, generating perspectives in the drug development eld to treat the infection caused by CHIKV and, potentially other alphaviruses.

nsP4-CHIKV puri cation and SEC-MALS analysis
The nsP4-CHIKV was expressed and puri ed using chromatography systems and the purity was con rmed by acrylamide gel electrophoresis (Supplementary Figure 1). The nsP4-CHIKV is formed by 492 amino acids and has a theoretical molecular mass of 54.54 kDa. This construction covers the full-length RNA-dependent RNA polymerase (RdRp) domain, responsible for the nsP4-CHIKV function and where the catalytic aspartic acid residues (Asp253 and Asp348) are located (20). The Asp348 is in the GDD motif, a highly conserved sequence of viral polymerases (20).
To determine the oligomeric state of nsP4-CHIKV in the working buffer solution, SEC-MALS was employed to estimate the accurate molecular mass (MM) of the protein without the need to use globular proteins as standards for mass determination (21). The SEC-MALS data showed a low polydispersity index and yielded a MM of (60 ± 1)kDafor nsP4-CHIKV ( Figure 1A). The proximity of the experimental value with the theoretical molecular mass of the protein suggests that the monomeric state is the most populated oligomer under the evaluated conditions. Evaluation of nsP4-CHIKV secondary structure pro le The CD spectroscopy was used the estimate the secondary structure content of nsP4-CHIKV in solution (22). The CD spectrum of nsP4-CHIKV is characteristic of an α-helical rich protein, with two negative minima at 208 and 222 nm and one positive maximum around 195 nm ( Figure 1B). This pro le corroborates with structural features described for viral polymerases and other polymerase structures solved experimentally (20).
The Dichroweb (23) was used in the quantitative analysis of the CD spectrum. In this way, the CDSSRT method (24) presented an NMRSD of 0.009 for the SP175 database and 0.012 for Sets 4 and 7, considered the most adequate t and the most representative of the secondary structure content of nsP4-CHIKV. Secondary structure content estimation from the CD spectrum con rmed the higher percentage of α-helix in the nsP4-CHIKV (60%). The remaining fractions were: 6.3% of sheets, 9.6% of turns and 24% of disordered structure. nsP4-CHIKV thermal stability pro le through CD spectroscopy The CD technique was also applied to study the thermal behavior of this protein when subjected to temperature variations (25). Due to its helical pro le, the folding transitions of nsP4-CHIKV was studied by monitoring the ellipticity at 222 nm as a function of the temperature variation (25). The results showed that the thermal denaturation process occurred cooperatively, presenting the transition from folded to unfolded states in a de ned way ( Figure 1C). From this result, the thermodynamic parameters of nsP4-CHIKV unfolding were obtained with the adjustment of a two-state equilibrium model to the experimental data. Changes in thermal capacity were not considered and the adjustment was made taking into account the linear changes in pre-and posttransition ellipticity, as a function of temperature variation (25). Thus, it was possible to obtain the T m and the apparent enthalpy variation (ΔH app ) for the same protein concentration at two different heating rates ( Table 1). The variation of the apparent entropy of the unfolding transition at each T m was calculated as ΔS app = ΔH app / T m , considering ΔG = 0 for each T m Table 1 Thermodynamic parameters of nsP4-CHIKV unfolding by CD spectroscopy. The melting temperature (T m ) and the apparent enthalpy change (ΔH app ) were determined by tting the CD data to a two-state equilibrium model (25). The apparent entropy change (ΔS app ) of the unfolding transition at T m was calculated as ΔS app = ΔH app / T m since ΔG = 0 at T m .

Differential Scanning Calorimetry (DSC)
The thermal behavior of nsP4-CHIKV was also evaluated using the calorimetric method DSC (27). The DSC experiments were performed at different protein concentrations and different heating rates. Figure  1D shows the temperature-dependence of the heat capacity pro le (C p -sample minus reference) of nsP4-CHIKV along with the instrumental buffer baseline, acquired with a scan rate of 10 o C/h. As a globular protein, the thermogram comprises three events: the pre-transition baseline, corresponding to the heat capacity of the native state (C pN ), the endothermic unfolding transition with a melting temperature (T m ) at approximately the maximum value of the C p curve, and the post-transition baseline, representing the C p of the denatured state (C pD ).
DSC experiments were performed at different scan rates to test its effect on the DSC pro le and the reversibility of the transitions. The temperature-and scan rate-dependence of the excess heat capacity pro le of nsP4-CHIKV are illustrated in Figure 1E. The nsP4-CHIKV undergoes an irreversible thermal denaturation since the second heating scans obtained at all scan rates used do not show any signal.
Moreover, the transition peak shows a clear scan rate dependence, characteristic of kinetically-dependent irreversible processes (28, 29). Indeed, all thermodynamic parameters associated with nsP4-CHIKV thermal denaturation exhibit a scan rate dependence ( Table 2). The dependences of the thermodynamic parameters on the heating rate are markedly non-linear (Supplementary Figure 2). This feature illustrates the non-equilibrium character of its denaturation processes. The dependence on the heating scan might also suggest that the rates of the conformational transitions are rather low compared with the heating rates.
The dependence of the T m on the scan rate was used to calculate the kinetic activation energy, E a , for the irreversible nsP4-CHIKV thermal denaturation. According to Sanchez-Ruiz et al. (29) the T m shifts induced by different heating scan rates, ν, for an irreversible two-state process can be modeled by the following equation: where A is the pre-exponential factor in the Arrhenius equation and R is the gas constant. Thus, by plotting against 1/T m , the apparent activation energy can be determined from the slope of the curve. The Arrhenius plot showing the scan rate-dependent changes in the T m is illustrated in Figure 1F, from which E a was determined as (110 ± 4) kcal/mol, similar to other proteins (30).

Evaluation of nsp4-CHIKV interaction with compounds
In the search for new compounds able to interact and inhibit the nsP4-CHIKV in solution, an initial experimental screening with a series of 12 compounds was performed using DSF (or ThermoFluor). The DSF can detect changes in the apparent T m of a protein in the presence and absence of ligands, resulting in an estimate of the thermal shift (ΔT m ) derived from these differences. This shift can be used as a primary indicator of complex formation (31).
For nsP4-CHIKV, the T m in the absence of compounds (only with DMSO) was 37.7 ± 0.4. Among the compounds, one called LabMol-309 ( Figure 2A) caused the highest thermal shift (~4°C). Thus, this data suggests the occurrence of interaction between LabMol-309 and nsP4-CHIKV and this compound was chosen to proceed with the other assays.
The interaction between LabMol-309 and nsP4-CHIKV was later analyzed by MST and solution-NMR. The MST data showed an a nity curve with the occurrence of well-de ned bound and unbound states ( Figure  2B). From that, the estimated dissociation constant (K D ± ∆K D ) for the interaction of nsP4-CHIKV with LabMol-309 was (6 ± 1) µM.
The interaction of LabMol-309 with nsP4-CHIKV was further evaluated by solution-NMR, monitoring the chemical shift perturbation (CSP). Figure 2C shows the spectra obtained for the compound LabMol-309 in the presence (in red) and absence of the protein (in blue). These spectra were superimposed, and the chemical shift differences were identi ed and mapped according to the respective positions of the proton resonances ( Figure 2C) previously identi ed in LabMol-309 assignment (Supplementary Figure 3).
Therefore, the detection of these chemical shifts perturbations is additional evidence for the interaction between nsP4-CHIKV and compound LabMol-309, reinforcing the results obtained through DSF and MST experiments The nsP4-CHIKV three-dimensional model and structural analysis A combined analysis of several biophysical techniques suggested that the nsP4-CHIKV is a monomeric αhelical rich protein capable of forming a complex with the compound LabMol-309 within a moderate dissociation constant (10 -6 M). The nsP4-CHIKV 3D structural model was obtained using AlphaFold2 (32). The model obtained showed a very high per-residue con dence score (pLDDT) for more than 90% of the covered sequence, and only for the N and C-terminal regions pLDDT were low ( Figure 3A). The fulllength nsP4-CHIKV structural model is illustrated in Figure 3.
In the nsP4-CHIKV structure, it was possible to identify the regions corresponding to the ngers (residues 93-315), palm (residues 316-502) and thumb (residues 503-611) domains, which are characteristics of viral polymerases ( Figure 3B). In the catalytic site, located in the palm domain, it is possible to identify the catalytic aspartic acid dyad (Asp371 and Asp466) separated by 7.5 Å. Besides, it is also possible to notice that nsP4-CHIKV contains an extra N-terminal domain (residues 1-92) forming a coiled coil. A similar motif is observed in the N-terminal region of the nsp8 from the SARS-CoV-2 RdRp complex, likely responsible for proper positioning of the nascent strand of RNA during polymerization.
In the model of nsP4-CHIKV, it is also possible to notice that the electrostatic surface potential surface of the active site cavity is remarkably positive, a signature feature of nucleic acids-interacting motifs ( Figure   3C). More than that, the analysis using ConSurf reveals that the catalytic site region and its surroundings are highly conserved ( Figure 3D). All this serves as corroboration for the robustness of the AlphaFold2 model of nsP4-CHIKV.

Molecular docking and molecular dynamics of LabMol-309 against nsP4-CHIKV
To investigate the binding mode of LabMol-309 at the nsP4-CHIKV, docking calculations were performed using the DockThor server. The docking results suggest that LabMol-309 binds to the nsP4-CHIKV active site, interacting with residues of the GDD catalytic triad (Asp466 and Asp467), with a docking score of -7.15 Kcal/mol -1 . LabMol-309 makes H-bonds with Glu369, Asp466, Asp467, Gly507, Arg573 and cation-π interactions with Lys295 residue (Figure 4). The nitrogen atom of the indole group and amine of the pyridine group of LabMol-309 make relevant interactions with Asp466 and Asp467, respectively.
Additionally, the indole group makes cation-π interaction with Lys295.
The structural stability of the structural model of the nsP4-CHIKV/LabMol-309 complex calculated by docking was evaluated using 100 ns molecular dynamic (MD) simulation. Figure 5A presents the values of root mean square deviation (RMSD) for the backbone atoms of the protein and non-hydrogen atoms of the ligand from the initial structure. It is possible to observe that RMSD values are stable after 5 ns of simulation and reach plateaus around 0.3 and 0.5 nm for the protein and ligand, respectively. Figure 5B shows that the number of contacts < 0.6 nm between nsP4-CHIKV and LabMol-309 do not drop down to zero throughout the MD simulation, indicating the ligand interacts with the protein all the time. The number of hydrogen bonds is stable throughout the MD simulations and presents an average value of three ( Figure 5C). An evaluation of the hydrogen bonds with signi cant percentage occupancy (< 5%, Supplementary

Inhibition of CHIKV replication by LabMol-039 through replication-based and viral infection assays
Although the data involving the complex formation between nsP4-CHIKV and LabMol-309 are solid, it still not possible to conclude that this compound would have inhibitory activity against this enzyme, or even against CHIKV-infection. The biochemical characterization of Alphavirus nsP4 polymerase activity has also been challenging because this protein cannot perform its function on its own (33). Different parts of nsP4 recognize the promoters for minus and plus strands. However, the binding requires the presence of the other non-structural proteins (nsP1, nsP2 e and nsP3) to form the replication complex and enable the de novo RNA synthesis (14,33,34). The interactions between these proteins within the complex during replication are limited and mainly uncharacterized (14). The main reason for this lack of information is that no high-resolution structure is available for the polymerase or the entire replication complex (33).
Given these limitations, it was not possible to evaluate the effect of the interaction with LabMol-309 directly on the enzymatic activity of puri ed nsP4-CHIKV. Therefore, the inhibitory activity of the compound LabMol-039 was evaluated through the replicon-based screenings in a dose-dependent manner to determine its effective and cytotoxic concentrations (EC 50  To con rm the antiviral activity of the LabMol-309, HEK-293T cells were infected with CHIKVnanoluciferase, a recombinant CHIKV that express the Nanoluciferase reporter, at a multiplicity of infection (MOI) of 0.1 in the presence or absence of the compound at 5 µM. Nanoluciferase activity levels, which is proportional to viral replication, were assessed 16 hours post-infection (adapted from (35, 36)).
Viability assays were performed in parallel and DMSO 0.1% was used as untreated control ( Figure 6C). As a result, these assays demonstrated that the LabMol-309 decreased CHIKV replication by 73.5% with cell viability of 79.4% (Table 4; Figure 6C). Altogether, the results obtained from the antiviral assays suggest that LabMol-309 is a potential molecule to be further evaluated as an inhibitor of CHIKV replication, even though it showed high toxicity to the replicon cells. This isolated effect observed only in the replicon cells can be explained by the differences in the cell lines (BHK-21 versus 293T) and incubation periods used in the antiviral activity experiments (48 h for replicon-based screenings compared to 16 h for the viral infection assays). Nevertheless, LabMol-309 could be further optimized to reduce its cytotoxicity and increase the selectivity index in cellbased antiviral assays.

Conclusions
Given the absence of speci c treatments for CHIKV infection, it is important to study, at the molecular level, the viral replication proteins and their use as targets for antivirals development. In this work, the polymerase nsP4 CHIKV was characterized through experimental and computational biophysical methods, generating unprecedented structural and dynamic information about this protein. A high prevalence of ordered helical secondary structures was observed by circular dichroism, which also showed that the nsP4-CHIKV unfolds under a cooperative transition during thermal denaturation. The thermal denaturation was further studied using differential scanning calorimetry, indicating a kinetically controlled process.
Considering nsP4-CHIKV as a promising target, a screening of compounds was performed through the DSF technique, and a compound designated LabMol-309 was identi ed. This biomolecular interaction was validated through MST, resulting in an interaction at a low micromolar range. The NMR was also used to con rm the interaction of LabMol-309 and nsP4-CHIKV, demonstrated the occurrence of chemical shifts perturbations and in silico molecular dockings and molecular dynamics simulations suggested the possible binding mode of this compound and its maintenance at the active site. Finally, the inhibitory activity of LabMol-309 was evaluated in cellular-based antiviral assays using a CHIKV replicon system and a reporter virus, demonstrating the inhibitory potential of this compound.
In summary, this study highlights biophysical features of nsP4-CHIKV, contributing to basic research on alphaviruses polymerase, and identi ed a new compound as a promising antiviral agent against CHIKV infection. These ndings could contribute to development of novel candidates targeting nsP4-CHIKV and support the progress in the eld of therapeutic strategies for CHIKV and other alphavirus infections.

CHIKV nsP4 cloning and overexpression
The coding region of nsP4-CHIKV (GenBank KP164572.1; PROT-ID AJY53709.1 -residues 118 to 611) was subcloned in the pET-SUMO vector using the LIC methodology (37). This construct encodes an nsP4-CHIKV with a N-terminal 6xHis-tag followed by a TEV protease cleavage site (ENLYFQ;GAM) and the nsP4-CHIKV puri cation nsP4-CHIKV was puri ed using an AKTA Puri er System (GE Healthcare). The rst step was a nity chromatography, using a HisTrap HP 5.0 mL column (GE Healthcare) pre-equilibrated with buffer A (50 mM Tris pH 8.0, 500 mM NaCl, 10% glycerol). The elution was performed using 50 mM Tris pH 8.0, 500 mM NaCl, 250 mM imidazole, 10% glycerol. The buffer was exchanged through dialysis to eliminate the imidazole excess. The 6xHis-tag-SUMO was cleaved by TEV protease during overnight incubation at 4°C.
A second a nity chromatography step was performed using the same system to collect the HisTag-less protein obtained after TEV treatment. A nal puri cation step was done using size-exclusion chromatography on an XK 26/1000 Superdex 75 column (GE Healthcare) pre-equilibrated in gel ltration buffer (50 mM Tris pH 8.0, 200 mM NaCl and 5% glycerol). The eluted fractions were collected and analyzed by SDS-PAGE to con rm their purity and mass spectrometry was performed to con rm the presence of nsP4-CHIKV. The nal protein sample was concentrated using 30 kDa MWCO centrifugal concentrators (Vivaspin, Sartorius). Protein concentrations were determined spectrophotometrically in a Nanodrop 1000 spectrophotometer, using the measured absorbances at 280 nm and the theoretical extinction coe cient of 36,495 M -1 cm -1 .

Size Exclusion Chromatography coupled with Multi-angle Light Scattering (SEC-MALS)
The oligomeric state of the puri ed nsP4-CHIKV was evaluated by size exclusion chromatography coupled with multi-angle light scattering (SEC-MALS) in running buffer composed of 50 mM Tris-HCl pH 8.0 and 200 mM NaCl. For that, 50 µL of puri ed nsP4-CHIKV at a concentration of 1.5 mg/mL was injected in a Waters 600 HPLC system (Waters) coupled in-line with a UV detector, a mini DAWN TREOS multi-angle light scattering apparatus (Wyatt Technology), a column Superdex 75 Increase 10/300 GL (GE Healthcare), and a refractive index detector Optilab T-rEX (Wyatt Technology). The light scattering detectors were normalized with bovine serum albumin (Sigma-Aldrich) and the ow rate used was 0.5 mL/min. The data were processed using ASTRA7 software (Wyatt Technology) with the following parameters: refractive index of 1.331, 0.890 cP for the viscosity of the solvent, and a refractive index increment of 0.1850 mL/g. Protein solutions were centrifuged for 10 minutes at 10,000 x g at a controlled temperature of 4°C immediately before use.

Circular Dichroism (CD)
Far UV-CD spectra (195-280 nm) were measured in a Jasco J-810 spectrometer (Jasco Corporation, Japan) equipped with a Peltier control system and using a quartz cell with a 1 mm pathlength. The spectra were recorded from 280 to 195 nm, with a scanning speed of 100 nm/min, a spectral bandwidth of 1 nm and a response time of 0.5 s. All the protein samples were in a nal concentration of 2.5 µM diluted in water. Spectral deconvolution was applied to estimate the secondary structure content using the DICHROWEB web server (23). Three different methods were combined with three different databases to improve the reliability of the results. The detailed analysis of the results generated by these combinations is provided in Supplementary material (Supplementary Table 1

Chemical shift perturbation
The LabMol-309 resonance assignment was performed using a Bruker Avance III 600 MHz. 1H-13C-HSQC, COSY and TOCSY were acquired at 298 K using 1 mM of LabMol-309 in D 2 O. The interaction between LabMol-309 and nsP4-CHIKV was studied using a Bruker Avance IIIHD 500 MHz in a solution of 20 mM ( 2 H) 11 -Tris/HCL pH 7.5, 200 mM NaCl and 250 µM of LabMol-309. One-dimensional 1 H spectra in the presence and absence of 3 µM NSP4-CHIKV were acquired, and the chemical shift difference was analyzed. The data processing and analysis were performed using TopSpin 4.09.

Molecular docking of nsP4-CHIKV and LabMol-309
The docking calculations were performed using the DockThor VS web (46, 47), focusing on the active binding site (Asp371 and Asp466 residues). The nsP4-CHIKV and LabMol-309 structures were prepared using the Protein Preparation Wizard tool (48) and LigPrep tool (49). The docking grid was centered at the active binding site; grid size 20 Å; and grid coordinates x, y and z of -27.84 Å, 12.89 Å and 28.25 Å, respectively. The search algorithm precision mode was set up in the standard con guration of genetic algorithm parameters, with the soft docking mode activated. The PLIP server (50) was used to analyze the protein-ligand patterns (hydrogen bonds, hydrophobic interaction, cation-π, π-stacking, water and salt bridge interactions). Poseview server (51,52) was used to generate 2D interaction diagram and VMD program was used to render the 3D images (53).

Molecular dynamic simulations
The initial positions of the nsP4-CHIKV-bound LabMol-309 for the molecular dynamic (MD) simulations were obtained by the molecular docking results and its topology parameterizations (Molid 814093) were obtained from the ATB server (54). The MD simulations were performed by using GROMACS package version 5.0.7 (55). The molecular system of the protein-ligand complex was modeled with the GROMOS54A7 force eld (56) and SPC water model (57) FBS with 500 µg/ml G418 (Sigma-Aldrich). The development of this replicon cell line will be described elsewhere. The CHIKV expressing nanoluciferase reporter (CHIKV-nanoluc) used for the antiviral assays is based on the CHIKV isolate LR2006OPY1 (East/Central/South African genotype) and was produced, rescued, and titrated as previously described (35,36 Declarations Figure 1 Biophysical characterization of nsP4-CHIKV. A) SEC-MALS analysis of nsP4-CHIKV in solution. The proximity of the estimated molecular weight of (60 ± 1) kDa with the molecular weight of the protein suggests that nsP4-CHIKV is mostly at a monomeric state under the evaluated conditions. B) Secondary structure pro le of nsP4-CHIKV by CD spectroscopy. The nsP4-CHIKV spectrum suggests the predominance of helical secondary structures. C) Thermal stability of nsP4-CHIKV probed by CD spectroscopy. The ellipticity at 222 nm as a function of temperature for nsP4-CHIKV was recorded at 0.5 o C/min (left) and 1.0 o C/min (right) and transformed to the protein unfolded fraction. Solid lines are best ts to the CD data using a two-state equilibrium model. The thermodynamic parameters of the protein unfolding transition are summarized in Table 1